Search

This paper presents tegdet, a new Python library for anomaly detection in unsupervised approaches. The input of the library is a univariate time series, representing observations of a given phenomenon. Then, tegdet identifies anomalous epochs, i.e., time intervals where the observations differ in a given percentile of a baseline distribution. Epochs are represented by time evolving graphs and the baseline distribution is given by the dissimilarities between a reference graph and the graphs of the epochs. Currently, the library implements 28 dissimilarity metrics, i.e., 28 different anomaly detection techniques, and its extensible design allows to easily introduce new ones. tegdet exposes a complete functionality to carry out the anomaly detection, through a straightforward designed API. Summarizing, to the best of our knowledge, tegdet is the first publicly available library, based on time evolving graphs, for anomaly detection in time series. Our experimentation shows promising results. For example, Clark and Divergence techniques can achieve an accuracy of 100%, while the time to build the model and predict lasts for few hundreds milliseconds.

Assessment of software nonfunctional properties (NFP) is an important problem in software development. In the context of model-driven development, an emerging approach for the analysis of different NFPs consists of the following steps: (a) to extend the software models with annotations describing the NFP of interest; (b) to transform automatically the annotated software model to the formalism chosen for NFP analysis; (c) to analyze the formal model using existing solvers; (d) to assess the software based on the results and give feedback to designers. Such a modeling→analysis→assessment approach can be applied to any software modeling language, be it general purpose or domain specific. In this paper, we focus on UML-based development and on the dependability NFP, which encompasses reliability, availability, safety, integrity, and maintainability. The paper presents the profile used to extend UML with dependability information, the model transformation to generate a DSPN formal model, and the assessment of the system properties based on the DSPN results.